Estimation of dynamic panel threshold model using Stata

Abstract

We develop a Stata command xthenreg to implement the first-differenced GMM estimation of the dynamic panel threshold model, which Seo and Shin (2016, Journal of Econometrics 195: 169-186) have proposed. Furthermore, We derive the asymptotic variance formula for a kink constrained GMM estimator of the dynamic threshold model and include an estimation algorithm. We also propose a fast bootstrap algorithm to implement the bootstrap for the linearity test. The use of the command is illustrated through a Monte Carlo simulation and an economic application.

Bank Credit and Short-Run Economic Growth: A Dynamic Threshold Panel Model for ASEAN Countries

“…These di erences could be due to the fact that the country sample is not exactly the same in our empirical investigation. 7 Our results are line with those of Botev et al (2019) in which they cannot con rm that nancial developments have a negative impact on growth beyond a given level of development using dynamic panel data threshold model which allows endogenous threshold variable and regressors (Seo and Shin, 2016;Seo et al, 2019).…”

“…This approach has several merits as recalled above, but controls only the endogeneity bias for some important regressors like initial growth in economic growth regressions. One way to circumvent this potential problem of endogeneity for the threshold variable is to follow the approach of Seo and Shin (2016) and Seo et al (2019). In particular, Seo and Shin (2016) develop a rst-di erenced estimator GMM, that allows both threshold variable and regressors to be endogenous.…”

Nonlinear impact of financial deepening on income inequality

“…Namely, the dynamic panel threshold model of Seo and Shin (2016) with a threshold in r −g 1+g is estimated 41 using the implementation by Seo et al (2019). 42 The requirement of balanced panel that is needed for the implementation by Seo et al (2019) restricts substantially the effective number of observations in the OECD and EU samples. Therefore, we additionally included the results with a sample of all countries at hand and not only those from EU and/or OECD.…”

Is there a nonlinear relationship between nonperforming loans and bank profitability? Evidence from dynamic panel threshold

“…Following Seo and Shin (2016) and Seo et al (2019), the empirical model for the threshold effect in the nonperforming loans–profitability nexus in the Nigerian banking industry is specified as where PROF , NONP , SIZE , and CAP are profitability, nonperforming loans, bank size, and capitalization, respectively. The details on the measurement of variables are provided in Table 2.…”

“…In particular, the philosophical stance of this paper differs from the long list of papers in the literature (Alandejani & Asutay, 2017; Amuakwa‐Mensah, Marbuah, & Ani‐Asamoah, 2017; Atoi, 2019; Gabriel et al, 2019; John, 2018; Kjosevski et al, 2019; Kure, Adigun & Okedigba, 2017; Kuzucu & Kuzucu, 2019; Laila, 2017; Rosenkranz & Lee, 2019; Us, 2018). To address this issue, the dynamic threshold model of Seo et al (2019) built on the GMM method is adopted. The method enables the policymakers to determine the threshold level of nonperforming loans that guarantee stability in the banking industry.…”